Homework 2

.docx
Explain the issues encountered when trying to have a representative sample The survey claims that 'The UAS is a nationally representative panel of U.S. households recruited through Address Based Sampling (ABS)' to emphasis the poll's trustworthiness. However, the term 'national representative' does not mean it draws on a representative sample. In fact, surveys could be adjusted and sampled with a purpose to represent the country on certain dimensions. Sampling methodologies can lead to bias. This can be observed in the UAS sampling procedure. Firstly, sampling in the UAS is carried out in batches. There are currently 20 batches, targeting either the U.S. population at large, or specific subsets of it, such as the population of Native Americans, California residents, and Los Angeles County residents. The subsets including California residents, and Los Angeles County residents do not cover all states of the US. The pollsters limit their polling citizens subset of California state, the resulting sample is not likely to be representative of the national population as they claimed. Second, most batches use a two-stage sample design, in which zip codes are drawn first, and then households are randomly drawn from the sampled zip codes. Nonetheless, they also use the algorithm to compute the unweighted distributions of specific demographic characteristics (e.g., sex, age, marital status and education). It then assigns to each zip code a non-zero probability of being drawn, which is an increasing function of the degree of "desirability" of the zip code. As we know a sample must be representative and random. Therefore, the process that controls the selection of people based on location and demographic characteristics should not applied to avoid bias. Explain how it is possible to overcome those problems In Pew Research Centre report, they found that a sample that based on demographically representative of the country (through use of quotas or weighting) did not predict accuracy. In other words, the survey looking to representative with respect to age, sex, etc. does not mean to estimates for other outcomes are accurate. They proposed an online polling method called "opt-in" surveys. This designation (for example SurveyMonkey, Morning Consult and The Harris Poll) covers a wide range of products, from sophisticated quota samples to large databases of consumers. With statistical adjustments, it seems better to reflect the national population. In this sampling method, respondents are not randomly selected but found wherever they can be found: recruited through memberships in loyalty programs, ads on websites, or just because they signed up to join a survey panel. These kinds of polls can help to reduce bias. Besides, weighting technique with the right variables also help to overcome challenges of polling procedure. Criticize polling methodology (explain its limits) A poll may label itself "nationally representative," but that's not a guarantee that its methodology is solid. One source of error is the use of samples that are not representative of the population as a consequence of an inappropriate methodology. Polling organizations have developed many weighting techniques to help overcome these deficiencies. Moreover,
to ensure accuracy, choosing the right variables for weighting is more important than choosing the right statistical method Secondly, poll samples might not be representative samples from a population due to a non- response bias. It is because the responses of those who agree to be selected might be significantly different from those who decline. Furthermore, survey results may be affected by response bias. It happens when the answers given by respondents do not reflect their true beliefs. It is because pollsters want to generate a certain result or please their clients; thus, they deliberately manipulate the outcomes. Respondents also try to influent the result of a poll, for example advocating a more extreme position than they actually hold, in order to boost their side of the argument or give rapid and ill-considered answers. References https://www.pewresearch.org/fact-tank/2020/08/05/key-things-to-know-about-election- polling-in-the-united-states/ https://www.pewresearch.org/methods/2019/11/19/a-field-guide-to-polling-election-2020- edition/ https://www.pewresearch.org/methods/2018/01/26/for-weighting-online-opt-in-samples- what-matters-most/ https://en.wikipedia.org/wiki/Opinion_poll#
Page1of 2
Uploaded by hongthuypr on coursehero.com